1. Movie Magic: a case study for strategy
implementation with analytics
Customer Segmentation, Personalization & Strategies
Conceptualized & Created by: Ashish Kumar Singh
2. The case
We would like you to prepare a short presentation on MovieMagic as outlined below.
MovieMagic offer two types of service, both of which are to be considered in this presentation:
MovieMagic by post – receive DVDs of your choice through the post
MovieMagic instant – stream films through the internet
MovieMagic have several databases:
1. Names and addresses for all their customers and when they joined
2. All films, DVDs and games rented (or streamed) by each customer and when
3. Subscription package held by each customer, and how this has changed over time
How should they use this data to treat their customers differently?
In order to grow further and compete they have identified that they need to take a more
structured and strategic approach; putting customer data at the core of their decisions.
How would you use this data to get an understanding of MovieMagic customers?
How would this understanding help MovieMagic in marketing to the customers?
What else, besides marketing could this information help MovieMagic with?
How could they use this data to personalise the service – for example, with the films that they
recommend to their customers?
3. OBJECTIVE,SCOPE & ASSUMPTIONS
Objective Scope Assumptions
For the • Pan India • Volume vs.
MovieMagic Geography Value based
Customer Base: coverage analysis
• Country • Customer data
• Create a
level mandatorily has
Customer analysis address
Segmentation • Micro level [location].
Framework limits to city* • Micro level
• Personalizatio • Data defined as the
n sufficiency lowest level
• Derive limited to where
Strategies mentioned heterogeneous
*Segmentation is possible where heterogeneity resources. patterns are
is observed. Level at which homogenous clusters will be obtained
will be disregarded and +1 level will be defined as base micro level. visible.
4. APPROACH
Context Approach and Strategy
MovieMagic offerings Data Architecture Model Development Deriving Strategies
need to be analyzed
with respect to usage − Databases are created − Classification is done – Defining customers via
patterns, customer using various sources: based on Volume vs. metrics in terms of
base and behavior, − Customer
Value concept. clusters created for
sales & product each branch.
− Technique used:
− Media – Personalization
portfolio and come up − Cluster analysis
− Usage,etc. exercise for each
with strategies − RFM Technique cluster.
backed by customer − Common linkages are
created using Primary − Clusters/Groups – Strategy derived for
understanding & keys,IDs created based on each segment:
personalization. independent metrics – Financial
− New data definitions are available for each of the
derived.eg.current – Marketing
categories.
inventory,turn-around- – Sales
− Prediction models
time,etc.
developed for each – Supply Chain.
branch of fishbone.
5. DATA ARCHITECTURE
Subscription Description
Custom Name Addres Email Joined *Custo Subscri Subscri
ID
er ID s Date mer ption ID ption
Type Date 1 Online Yearly
Subscription
Customer Database 2 Offline Yearly
DB
Custom Media Date of Period Quantity Price
er ID ID Rent 3 Online Monthly
4 Offline Monthly
Rent Logbook DB
Cat. ID Descr. Class
Media Media Type Categor Invento Date Rented 1 Genre Movie
ID Name y ID ry (Q) Time Quantit
2 Starcast Movie
Media Category
Stamp y
3 Released Movie
[Movie/ Movie/
DB
Game Game 4 Languag Movie
Name] Media Database e
5 PC Game
6 PS Game
*Customer Type:Subscriber Online,Subscriber Offline,Repeater Online,Repeater Offline,One 7 Genre Game
Time Online,One Time
6. METHODOLOGIES:CLUSTERING/PREDICT
Segmentation
Offline Online
Rent Subscription Subscription Non-Subscribers
OneTime Repeaters Yearly/Monthly Yearly/Monthly OneTime Repeaters
Cluster Function Cluster Cluster Function of: Cluster Function of: Cluster Cluster
of: Function of: -DatetimeStamp -DatetimeStamp Function of: Function of:
-DatetimeStamp
- -Geography -Geography - -
-Geography
-Media Category
DatetimeStam -Media Category -Media Category DatetimeStamp Datetimestamp
-Media Type p -MediaType -MediaType -Geography -Geography
-No. of times -Subscription type -Subscription type - -
rented. -No. of times -No. of times MediaCategory MediaCategory
-Geography subscribed subscribed MediaType -MediaType
- -No. of times -No. of times -Browsing -No. of times
MediaCategory discontinued Discontinued history bought
-MediaType -Browsing history -Browsing
history
For each of the Fishbone branch>>Subset of data obtained>Clustering/RFM Technique Used>Model
7. IMPLEMENTATION
Sales Strategy Financial Strategy Marketing Strategy Supply Chain Strategy
USAGE PATTERN: DISCOUNTING: GEO SPECIFIC STOCKOUT/BACKLOG:
Discounting to ADVERTISING: More predict inventory to avoid
Within X days of
clusters of users stockouts,calculate
release, Y% of extra advertising in less
based on profit Adjusted Turn Around
streaming over base generation, less penetrating areas wrt Time based on
and thereafter Z% of penetration, usage index, Consumption Pattern for
extra rent over base opportunity index.Like competitive scenarios. each branch of Fishbone.
value after X days. Hike Prices when TARGET RED FLAGS:Flagging
demand more in
COMBO PACKS: MARKETING based Users which are probable
streamline for a
Creating **combos of on usage pattern & unsubscriber/
period,pattern of
DVDs to push sales of specific demands, discontinuation of usage
usage.
Customer Lifetime based on patterns of
Slow Mover
SUPPLIER subscription packages
DVDs,Contra Value.
NEGOTIATION: they used.
Sales[Demanding Based on usage
clubbed with non- DISTRIBUTION
pattern, demand NETWORK: local network
demanding] forecasting, days of at more demanding
payable outstanding areas.
can be negotiated with
Personalization: Creating suppliers of users
the portfolio at Individual level and implementing above
strategies.
**DVD Types:1 movie/game pack, N in 1 pack,Combo Packs,Star Packs,Vintage packs,etc.